人工免疫粒子滤波算法估计电动汽车电池SOC

准确预测电池的荷电状态（SOC）对纯电动汽车的安全可靠的运行具有重要意义.标准的粒子滤波算法对锂离子动力电池的非线性特征有一定的适应性，能够对电池的 SOC做出估计.但是在标准粒子滤波运算过程中普遍存在粒子退化现象，导致算法效率和预测精度降低.因此，本文提出一种新的人工免疫粒子滤波算法，将人工免疫算法的原理引入标准粒子滤波算法的粒子更新过程中，对锂离子动力电池SOC的估计进行优化，以提高SOC估计的准确性.利用北京市实际运营的纯电动汽车电池数据，对所提出的电池SOC算法进行实证研究.实验结果表明，相对于标准粒子滤波算法，人工免疫粒子滤波算法能够增加粒子的多样性，具有更好的SOC预测精度和有效性.

Estimation for SOC of PEV Battery Based on Artificial Immune Particle Filter

Accurate state- of- charge (SOC) estimation of batteries is important for the development of electric vehicles. However, a common problem with the particle filter is the degeneracy phenomenon, resulting in low efficiency and the estimation accuracy. Therefore, an artificial immune particle algorithm is proposed to optimize the estimation of SOC in this paper. Based on the battery data of pure electric vehicles (PEV) running in Beijing, This paper makes comparing experiment for SOC estimation. The experiment result shows that, artificial immune particle filter algorithm has better SOC estimation accuracy than standard particle filter algorithm.